publisher colophon

69Chapter 3

Citizen Security and Democracy

Citizen security has become a crisis in Latin America, primarily by feeding off the weaknesses of democracy. Frustration over crime and disorder has led to the biggest public protests since the regional transition to democracy in the 1980s and has helped to trigger the collapse of six constitutional governments since 2000. Presidents were forced out in Argentina, Bolivia (twice), Ecuador, Honduras, and Venezuela (temporarily). One United Nations report on eighteen Latin American countries found that a majority of people would support an authoritarian regime if it solved their country’s problems, of which crime is a principal one.1

Leading up to and beyond such attitudes and breakdowns, however, is crime’s corrosion of everyday life. Amid political and economic uncertainty, when the malleability of institutions and the swings in public opinion are at their highest, insecurity has been both a constant and a constantly changing feature of contemporary Latin America. In particular, as this chapter asserts, it has permeated three core areas of governance in the region: (1) power relations within state and society, (2) geographic differences among regions, and (3) civil society in each country. Citizen security has profoundly changed all three areas, and in ways that complicate democratic governance. A closer look at those areas, which this chapter undertakes, can help explain the persistence of obstacles to reform, even amid Latin America’s changing socioeconomic and political conditions.

Most generally, democracy can be defined as a regime that structures and channels citizens’ control over their government. The many principles and practices cited as integral to a democracy, from universal suffrage to the subordination of the military to civilian control, can be grouped into sets. One set is the balance of power, in which the branches of government—normally, the executive, legislative, and judicial—have clearly delineated powers that are counterweights to each other. A second set of principles constitutes the rule of law in which civil and political rights are upheld, primarily through an independent, functioning judicial system that is accessible to all citizens. Building on the first two sets, the third is a state able to implement policies in response to societal needs throughout its territory. All of the specific goals within these sets are carried out using a range of instruments such as written constitutions, oversight commissions, and electoral rules. Because of their importance, a transition to democracy can thus be regarded as “the process of establishing, strengthening, or extending” these instruments (Garretón Merino 1995).

In practice, however, these instruments are impaired. Constitutions are undermined by decrees, elections by party clientalism, and civil rights by politicized judges. Together, such impairments can stall a transition from authoritarianism to democracy.2 Recognizing this slow progress, many studies assess a transitioning regime by the degree to which it falls short of standards associated with a stable and consolidated democracy. But instead of examining what characteristics the Latin American democracies have not attained, this book looks at those they have attained. As this chapter argues, three areas of democratic life—power, geography, and civil society—reveal how fragmented these democracies are. The accumulation and use of power, the geographic disparities among regions and sectors, and the interactions and development of civil society all show how the principles of democracy splinter as they open up and expand. The obstacles to citizen security identified in the previous chapter incubate in these three areas, with minimal impact until they accumulate and together push consolidation off course. For example, pockets of disillusionment with law enforcement can erupt into political rallies that shape the public agenda, while informal power relations within the police can block reforms by legislators who are unaware of those power relations. In short, like other politically charged issues, citizen security reveals the fractured forms of democracy into which Latin America has settled over the last two decades.


Democratization opens up channels of power, but not always in ways that advance or augment democracy. Security forces, charged with the state’s physical might, are one of those channels. As Dahl (1998, 149) affirms, “perhaps the most dangerous internal threat to democracy comes from leaders who have access to the major means of physical coercion: the military and the police.” One of the achievements of Latin America democracy is reversing or minimizing that long historical tradition by subjecting security agencies to the rule of law rather than the interests of rulers. However, within the complexities of modern society the continuation of that achievement means more than simply subduing the police. It is not just the level of these agencies’ power that shapes democracy, this book argues, but its character. And increasingly that character is one of uncertainty. The manifestations of power that are dangerous to democracy are not just reversions to authoritarianism, but failures to stabilize into coherent forms and predictable practices. In a democracy, above all, the state’s power to control citizen security is accumulated and expressed not by force but by accessibility and consistency. It is not about what the state imposes on citizens, but what kind of information and cooperation it receives from them and what kind of service it provides for them. For example, acknowledging in August 2008 that his crime crackdown and planned structural reorganization were not working, Mexican president Felipe Calderón issued a more modest appeal for citizens to participate in crime prevention and report crime more often. Focusing on three relationships of power—those between police and society, those within the state network, and those among officials within each agency—this section discusses why the fragmented and uneven nature of power in democratic Latin America make advances in citizen security unlikely.

Police and Society

Expression of power between actors, as many scholars point out, is a relationship rather than simply a form of control. In a formulation by Bachrach and Baratz (1970, 21–23), A has power over B when B bows to A’s wishes, and when A can threaten to invoke sanctions that B regards as deprivations. The police usually have such power, and they make full use of it. But because for many citizens the police are the primary form of state contact for many of their needs, and because the police need citizens to prevent and solve crimes, such exercise of power is tempered by a mutual incentive to cooperate in many types of interactions. Thus the police often hold back on uses of power such as arbitrary arrests, and citizens refrain from legal actions. But these clear lines of engagement, both positive and negative, are being blurred in ways characterized by their lack of predictability. In Latin America the public is increasingly seeing policing as not simply discriminatory or deprivational, as in the past, but also illegitimate and unpredictable. In most places it is anyone’s guess whether officers will in fact try to solve a robbery, whether a police presence calms or inflames a domestic dispute, and whether a routine stop will mean a shakedown. Connections between police and organized crime, private security, and local elites make relations even more uncertain with citizens, who see officers as placing the interests of those entities above theirs. As Marenin (1996) stresses, the police maintain not only order, but a particular order that privileges certain groups, and often with an “erosion of the capacity and willingness of state agents themselves to abide by the rule of law” (Koonings and Kruijt 2004, 2). As a result, people usually do what they can to avoid the police, even when they are crime victims or witnesses. In situations in which police behavior is grudgingly accepted as legitimate, people still want to cut it off before it leads into the even more uncertain zone of criminal justice, which can trap crime perpetrators and victims indefinitely. When charges of abuse are involved, the improbability of discipline against the offending officer, as chapter 2 discussed, only adds to the tension and arbitrary nature of police-society relations.

The State Network

Along with police-society relations, some of power’s most vital arteries connect the network of agencies that make up the modern state. For these interstate channels to work democratically, each agency must be both accountable to laws and empowered to carry out policies (see Geddes 1994). But powerful entities that do not abide by the rules, as many scholars have discussed, have played outsized roles in the formation and character of the modern nation-state (Hobsbawm 1985; Kennedy 1987; Tilly 2003). The military has been predominant among such “reserve domains of power.” In regions like Latin America where most armed threats have been internal rather than external, the police have been predominant as well. In every sub-region of Latin America they were central actors in the extended civil wars following independence in the 1800s. As the fast economic growth that began in the late 1800s put a premium on minimizing the social disruption that comes with it, the police became a primary building block of the modern state. As those states became increasingly authoritarian in the 1900s, that role gave police agencies greater political and infrastructural power, freeing them from many of the rules and constraints to which other state agencies were bound. But unlike the military, which was usually purged and restructured during democratic transitions at the end of the 1900s, the police agencies that were also part of the repressive apparatus remained largely untouched by fledgling governments unable to take on yet another challenge. With political and economic uncertainty continuing in most countries—and with the military no longer available to contain it—the police then emerged as their countries’ primary security force and guardians of the state’s monopoly of legitimate violence. Through their physical power in society, institutional power in the state, and political power in the government, the police have made themselves indispensable on the range of security issues, from neighborhood safety to the mass political upheavals that have threatened governments throughout the region.

Once the underlying strengths, allegiances, and interests of state agencies come into greater focus, so too do the ways in which power is transmitted. A good example is information, which, like the law, is supposed to facilitate free expression and check abuses in a democracy. But just as information is hidden in the free market, it is also hidden in the public arena. On one level the police’s de facto control over evidence, witnesses, and other information makes judges, fiscales, and public defenders dependent on them. From the technical mastery of advanced weapons to politicians’ dirty laundry, information is also used by police officers (as well as by criminal organizations) to shape criminal policy. Such informal controls complicate the accountability mechanisms described in chapter 1, as well as the citizen-police distrust explored in chapter 2. Policing is also affected when the technology to track a certain crime statistic leads officials to emphasize that particular crime over others. By contrast, the lack of information is experienced most acutely by criminal defendants, whose basic right “to receive an explanation of what has been done” (Sklar 1999, 53) is routinely violated as they wait years in inhumanely overcrowded prisons without regular access to a lawyer.

Access to local power figures and civilian criminal justice officials, or foreign agencies, is also a form of power currency among officers. Many police agencies wield their agreements with other governments to gain a political upper hand. Many programs become too closely associated with those officials who guide them through the decision-making mill, which increases the likelihood that the program will end when those officials end their tenure. The United States, regional reform’s biggest funder, has shown how aid may also deepen both donor and recipient divisions. During the Cold War,3 American training of foreign police reflected differences between the civilian-oriented approach of the State Department and the Office of Public Safety in the U.S. Agency for International Development (USAID), on the one hand, and the more militarized approach of the Defense Department and its regional counterparts, on the other. Such tensions were tamped down by Section 660, the 1974 amendment to the Foreign Assistance Act prohibiting assistance to foreign police. But the differing post–Cold War priorities in Latin America among the main U.S. agencies working on security—State Department, Justice Department, Defense Department, and USAID4—have revived these contradictions, such as between offering training on new due process rights and funding new police units that ignore them. Drug trafficking in the Andes and youth gangs in Honduras, as this book discusses, are two issues over which such clashes have occurred, weakening citizen policy in recipient countries.

Such uses of outside power raise larger questions about the nature of democracy and the state. Just as judiciaries under the thumb of an authoritarian elite may gain more independence in dealing with certain issues than those in a transitional democracy trying to maneuver among competing elites, a stable, narco-dominated fiscalía general can be more “independent” in dealing with low-level cases. Even in internal state relations “the line between state and society is not the perimeter of an intrinsic entity, which can be thought of as a freestanding object or actor,” observes Mitchell (1991, 90). “It is a line drawn internally, within the network of institutional mechanisms through which a certain social and political order is maintained” (emphasis in original). When such penetration sets the basic practices of a state agency, it can alter the basic actions and standards of democracy. Thus when organized crime controls a police unit, who is actually providing security? When agencies are more respectful of certain legal standards to avoid attention to their corrupt activities, what happens to the rule of law? States riddled with illegal activities tend to be considered “weak or in crisis” (Van Reenen 2004, 33–34), but in fact such activities may become integrated into what become some of the more stable realms of state functioning. It is not just pieces of state power that are being seized, but the nature of state power itself. The penetration by organized crime of Central America’s criminal justice bodies, as well as the other cases discussed in this book, thus complicate the concept of “authoritarian enclaves” (Garretón 2002, 85) and assumptions about the complementary advancement of democracy, state strength, and the rule of law.

State Agencies

A third significant dimension of power is the people employed by state agencies. Like the police, the matrix of units and officials who form large bureaucracies runs on various currents of power. Many of these currents become visible in the realms of change discussed in chapter 2, such as the ongoing competition for promotion among officers or for influence in a newly decentralized police force. To survive and thrive in such environments, officials must draw on different sources of individual power. For the police, four of the most important types of such power are hierarchical, evaluative, innovative, and reputational. Hierarchical power, based on rank and position, begets obedience of an officer’s directives and commands. It is usually accompanied by evaluative power, which is the ability to distribute or withhold resources, promotions, raises, and other objects of value. Innovative power allows the bearer to design new programs or policies—and then get them implemented. Closely related is reputational power, which is based on personal qualities such as charisma and respect by colleagues. In many institutions the biggest drag on effectiveness is when innovative power and reputational power do not allow officials to move up the ranks to acquire hierarchical and evaluative power. In a police force the combination of a militarized history, the centrality of seniority, and the distortions affecting “high” organizations with too many ranks makes such movement particularly sticky. Moreover, as in other agencies, power is awarded to and well guarded by police veterans who entered into the system early and assiduously worked their way up. Top-heavy structures crowd out those below, slowing the efforts of mid-ranking officials to develop new programs, transmit information from the street to headquarters, or reward patrol officers’ innovation. When officials manage to break through such barriers, automatic rotations can stop the accumulation of knowledge and relationships that help uncover the causes of crime and help turn reputational power into hierarchical power—that is, the management problems described in chapter 2 can often be traced back to the power relations among officers.

These three dimensions of power—state-society, interstate, and intrainstitutional—are evident throughout a democratic regime. They directly and indirectly shape many of its basic practices and standards, including accountability, a key principle discussed in the first two chapters. After a history of authoritarianism, as the scholarship emphasizes, state institutions that abide by the law are a linchpin for nearly every democratic standard. If a democratic regime is consolidated when all the major actors “become habituated to the resolution of conflict within the specific laws, procedures and institutions sanctioned by the new democratic process” (Linz and Stepan 1996, 8), then mechanisms to advance and monitor such habituation are indeed central to democracy. In the many definitions of accountability, two general dimensions are cited as essential to it. The first is explanation, transparency, and answerability—the power to know what is happening and why. The second is judgment and sanction—the ability to punish those who violate laws.5 Accountability agencies, such as those discussed in chapter 1, have become increasingly able to reach the first goal and, under favorable conditions and cooperative courts, the second one as well. Ombudsmen in the Andean region, for example, have successfully addressed cases involving abuses such as price fixing, violence against army cadets, and discrimination against indigenous persons. But they still struggle to alter the conditions that contributed to such abuses, such as loose regulations over companies, military bodies, and local governments. In many instances, though, accountability agencies can be thrust into the center of political controversies that force governments to strengthen them. For example, the 2010 Gulf of Mexico oil spill led to demands to strengthen the U.S. Interior Department’s Minerals Management Service, which oversees oil extraction practices. In such cases, accountability agencies will be allowed to become sufficiently independent and institutionalized to help bring about “horizontal accountability” among state agencies (O’Donnell 2003) or between the state and the private sector.

In most countries, however, accountability is under threat by the widespread belief that it hinders the decision making necessary to overcome persistent economic and political problems. Governments “immobilized by oversight mechanisms,” as Mainwaring (2003, 4) explains, may be “perceived as indecisive, ineffective, or inept.” Unconstitutional activities, from individual corruption to federal intervention, are tolerated if the public views them as necessary to—or as a reasonable price to pay for—decisive action. With crime now one of the public’s biggest concerns, support for crackdowns that may even involve extrajudicial killing is clashing with laws and agencies formed during democratization to hold the police accountable to basic standards. In a poll in Maracaibo, Venezuela’s second-largest city, 91 percent of respondents did not “feel secure” with the police, but an astonishing 47 percent favored police killing of delinquents (Gabaldón and Bettiol 1988). Stated differently, outright fear of a state agency does not necessarily mean that citizens want it to be held accountable. Even pro-reform governments such as Chile’s have had to be on good terms with public safety agencies, forcing it into a “trade-off between promoting citizens’ rights and defending the police from allegations of misbehavior” (Fuentes 2006b, 155–56). So despite the fact that accountability agencies around the region have made important advances in gathering knowledge and administering punishment, they have yet to realize a more substantive third dimension—uprooting the sources of unaccountable action.

Even though the biggest underlying reason for poor accountability is an acceptance of its absence, shifts in public opinion can alter unaccountable exercises of power. Peruzzotti and Smulovitz (2006, 4) emphasize “the growth of alternative forms of political control that rely on citizens’ actions and organizations” and that can set accountability mechanisms into gear through social mobilizations, legal claims, and media exposure. However, because of the institutional and political insularity of the police, such strategies are less effective in law enforcement. As Schedler (1999, 25) has observed, “Civil association may represent strong ‘factual power,’” but citizen groups “can never match the state’s monopoly of legitimate physical violence or its privileged position as the source of law.” Weak opposition to get-tough policies in most countries further enhances the government’s ability to hide the abuses arising from them. Support for a policy of mano dura by a wide cross section of society provides additional cover by neutralizing the argument that it targets a particular class or geographic area. Accountability is certainly compatible with tough action because it can help make the police more efficient, but mano dura proponents usually cast them as inherently incompatible, as discussed in chapter 1, thereby neutralizing the contribution of accountability to traditional policing. Such manipulation of the levers of power between state and society and within the state clearly maligns democracy’s principle of accountability when it comes to citizen security.

The communities suffering the most from crime are the ones with the least power to work with or challenge the police—that is, they experience the greatest economic pressures, ethnic tensions, residential turnover, and suspicion of both police officers and anyone who cooperates with them. As numerous authors describe (see Holston and Caldeira 1998), the decision by police officers to use force often depends on the citizen’s demeanor and perceived power. And the vast majority of these discretionary judgments are not challenged or overseen through accountability mechanisms, demonstrating the difficulty of reversing trends beyond the individual cases that are documented. Legal and constitutional analyses often miss most of what happens beneath the surface of public scandals and indictments, such as detention beyond the legal time limit, forced confessions, and evidence tampering. Such actions are particularly difficult to monitor because they happen on the street and to citizens with little access to the means for challenging them. Similar to the invisible matter that makes up the bulk of the universe, the sum of daily actions gives shape to the state and its power relations with society.

A closer look at ombudsmen (defensorías del pueblo), the main accountability agents in contemporary Latin American democracies, will highlight these barriers. Around the region ombudsmen are empowered to investigate abuses by any state agency, most often by directing the complaints to the appropriate channels. The high volume of complaints filed about police abuse is testament to the need for accountability on security. For example, well over half of all grievances to El Salvador’s human rights commissioner involved the National Civilian Police (Chinchilla 2003, 215), and in Peru the police received the third highest number of citizen complaints.6 In 2004 Nicaragua’s Office of Internal Affairs received 1,113 reports that led to the investigation of 639 police officials (Programa Centroamericana de la Federación Luterana Mundial 2005). And in the city of São Paulo the police auditor received 16,000 complaints of police abuse in just the first two and a half years of that agency’s existence (Miller 2002, 16). Because of their importance to such issues, along with the great strides they have made on many of them, defensorías enjoy high levels of citizen support around the region. Most governments also express rhetorical support for them, which is, at least for Costa Rica, considered to be genuine.7 But most defensorías find themselves constricted by pressure from the executive branch, a dependence on state funds, a lack of patronage networks, resistance by the bureaucracies they try to pry open, and a mandate that stretches them too thin. As a result, defensorías tend to be most effective against weak governments, such as in Peru in the late 1990s, or responding with popular backing on highly charged issues such as rigged elections or utility price hikes. As Schedler (1999, 23) points out, “an accounting party, if it is to be serious, cannot stand on equal footing with the accountable party. At least in its sphere of competence, it must be even more powerful.” Most Latin American accountability agencies lack the resources and autonomy to be on such footing in most areas of state power.

Such limits on ombudsmen have prompted most countries and provinces to establish agencies with the sole purpose of monitoring the police (see table 3.1). Although varying in size and legal reach, almost all of these agencies have the authority to investigate police wrongdoing, publicize abuses, and prod judicial and executive agencies to action. In most countries they have focused primarily on internal management (usually corruption) and civil rights (abuses from arbitrary detention to extrajudicial killing). To minimize politicized interference in such sensitive matters, most internal affairs agencies are bolstered through measures such as legal and financial autonomy and mixed civilian and police staffs.

Internal affairs units in Peru, Argentina, and other countries have made major advances, but the majority have been stalled or rolled back by use of the state’s financial, institutional, and political powers. For example, starved of resources and inundated with minor complaints that prevent investigation of serious ones, many units cannot carry out investigations or are forced to allow the police to investigate themselves. Internal affairs units in countries such as Paraguay, Nicaragua, and Panama have not had much control over police abuses. Even Peru’s internal affairs structure, one of the most extensive in the region (see Basombrío Iglesias 2006), has been weakened by resistance from the police, overlap with the inspector general, and the revolving door of six interior ministers in the five years after Alberto Fujimori’s regime ended in 2000 (Gobierno del Perú 2002). Like their defensoría counterparts, these internal affairs units are weaker than the institutions they try to hold accountable. As a result, many police are able to use edicts, “social order” statutes, and other regulations to build a legal firewall against them. Internal affairs units are also circumvented, ignored, or second-guessed by the other bodies with which they work in investigating malfeasance. Security officials often act less than expeditiously when sending material on specific cases, and the judges and executive officials who try and sanction charges of police abuse are often more lenient than accountability agencies.

Table 3.1 Police Accountability Mechanisms in Latin America


Police accountability agencies and procedures


Inspección de Seguridad (most provinces)


Dirección Nacional de Responsibilidad Profesional; Tribunal Disciplinario Superior


Ouvidoria da Polícia (fourteen states)


Administrative procedures; military courts


Defensoría del pueblo (national ombudsman)

Costa Rica

Consejo de Personal

Dominican Republic

Comisión Nacional de Derechos Humanos; Instituto de Dignidad Humana—Policía Nacional


Defensoría del pueblo

El Salvador

Dirección General de Inspección—Policía Nacional Civil


Comisión de Derechos Humanos de Guatemala; Inspectoría General—Policía Nacional Civil (IGPNC), consisting of the Oficina de Responsabilidad Profesional (ORP, investigates charges), the Oficina de Derechos Humanos (ODH), and the Régimen Disciplinario (REDIS)


Unidad de Asuntos Internos; Consejo Nacional de Seguridad Interior


Varies by state; federal-level Consejo Nacional de Seguridad Pública


Dirección de Asuntos Internos


Unidad de Asuntos Internos


Unidad de Asuntos Internos


Police ombudsman; Dirección de Asuntos Internos (Interior Ministry); defensoría del pueblo (national human rights ombudsman)


Inspección Nacional de Policía


Defensoría del pueblo (national); Dirección del Inspector (local and state)

Source: Author.
aIn federal countries, differences among provinces or states are wide.

Limits on time, information, and resources on many fronts force internal affairs units to respond to rather than prevent abuse, just as traditional policing does with crime. Such ineffectiveness is evident in administering discipline, a major instrument of state authority often exercised through hierarchical power. In most countries charges against police officers are handled in one of three ways: (1) through internal police procedures, usually by the accused officer’s superiors or special courts; (2) through the defensoría del pueblo or human rights commission; or (3) through the prosecutor’s office (fiscalía), either directly or via the police or defensoría. Serious cases sent through all three channels may end up in criminal court. But the tendency of the police to close ranks behind accused colleagues slows action by already overwhelmed administrative, defensoría, and judicial officials. Venezuela’s police officers are rarely convicted of abuse, or they receive sentences far shorter than civilians convicted of the same crime (PROVEA 1996). Of the three hundred specific cases of police abuse documented by rights groups in Venezuela between 2000 and 2005, only thirteen led to formal accusations and two to sentences (La Red de Apoyo 2006, 27).

Because of such conditions, in many countries charges are processed internally. Although it may make administrative sense, this approach can keep much information away from internal affairs units and civilians. For example, Ecuador’s tripartite system of tribunals, district courts, and National Court of Police Justice used to try accused officers is officially independent, but its impartiality is open to question because many of its judges are retired police officers. Limited in their controls, many governments opt for purges to try to sweep away all problematic officers in one fell, highly publicized swoop. Not only can such purges create a mass of angry former officials motivated to create trouble, as in Argentina, but they also usually fail to eliminate the uses of power that allow police abuses to survive those personnel changes, such as in Bolivia.

All three of the countries featured in the case studies in this book established police accountability units in the late 1990s. Each of their experiences reveals how, even amid dissimilar politics and similar problems, accountability can rarely venture beyond a regime’s political tolerance level. Public pressures for action by the state, amid a lack of confidence in it, justifies officials’ self-appointed prerogative to utilize certain agencies unhindered. Amid the institutional changes and mismanagement discussed in chapter 2, such selectivity limits oversight during the life span of state policies, from initial formulation to excesses in implementation. It is only in such fragmented states that repeated actions on specific abuses have just temporary impacts. Accountability agencies may have notched up many victories, but they have yet to win the war. Because of tolerance of abusive policing, accountability agencies find it difficult to attain not only answerability and punishment, but also the more elusive and arguably more important goal of addressing and reversing the causes of abuse.


Although one of the most important attributes of a regime is whether it “can exercise authority throughout the society” (Weiner and Huntington 1994), the disparity in the quality of basic security provided to different regions within nearly every Latin American country is ever-widening. These differences entrench unequal power relations, exploitative state actions, and societal mistrust. As one of the most overlooked dimensions of criminal justice, geography thus amplifies the obstacles to citizen security reform. To explain how, this section discusses three overlapping layers of geography—decentralized state structure, urbanization, and privatization—and how they shape policing and criminal policy.

Decentralized State Structure

A primary cause of the geographic differences among regions is the unprecedented decentralization of Latin American states. Like many other services, citizen security has been decentralized to the regional and local levels, resulting in a record number of police forces. But decentralization may turn out to be not just more inefficient than centralization, as chapter 2 discusses, but more undemocratic as well. In most countries subnational governments do not necessarily allow more participation, accountability, or rights. Instead, decentralization can breathe life into the local prejudices and political clientalism that were held in check by national standards. Even if such patterns do not metastasize into “decentralized despotisms” (Mamdani 1996), their impacts on daily life can be wide-ranging and deep, as evidenced by the lack of protection for civil liberties in many rural areas of Argentina, Brazil, and Venezuela. Such subnational power has been latent throughout Latin America’s history. Because the region’s populist and authoritarian regimes stopped or carefully controlled change through the 1900s, local powers were able to survive through the century’s otherwise transformative changes such as urbanization, economic modernization, and the formation of mass parties. Many current efforts to rationalize or renegotiate police controls between central and regional governments are now being complicated by these local powers. For example, recalcitrant provinces of federal states such as Argentina are resisting change; Mexico City and other areas are attempting to “impose authority on a police system … fragmented and out of control” (Davis 2009); and in Bolivia local powers are posing threats to the integrity of the state itself.

Amid economic inequalities, decentralization can also mean wide differences in the quality of state services, including police, among regions and cities. In every country per capita urban crime rates are far higher than the national average. Ecuador’s National Institute of Statistics and Census (Instituto Nacional de Estadísticas y Censos de Ecuador, INEC) reports that the murder rate in the country’s cities is about three and a half times the rate in rural areas, for example, and in Bolivia 80 percent of murders occur in its three urban departments. Although Venezuela’s national homicide rate was 48 per 100,000 persons in 2007, Caracas’s was between 130 and 166.8 Eager to be relieved of this urban burden, provinces without large metropolises tend to support decentralization. The smaller and more self-sufficient provinces benefit from decentralization, but poor ones usually end up worse off. In Nicaragua, which has one of Latin America’s more equitable distributions of income, the high-poverty rural areas representing 16 percent of the population have only four hundred officers, which is one for every 2,188 persons, whereas the rest of the country has one for every 710 persons (Bautista Lara 2006, 114). Moreover, because citizen security depends on agencies beyond the police, the proportionately greater presence of some agencies determines not only the level but the kind of protection. For example, a scarcity of public defenders or social services can diminish certain components of citizen security, such as prevention and civil rights, even more than a shortage of police. This dimension of inequality accentuates the state’s fragmentation in part by propagating the obstacles to citizen security throughout a country.


Decentralization has an even greater impact within urban areas, where violent crime is concentrated in the neighborhoods least able to afford quality policing but most in need of it. Many scholars in the fields of human ecology and social urban geography, postulating “an intimate congruity between the social order and physical space,” have documented this relationship between demographics and crime (Georges 1978, 3). For example, Burgess (1925) and Shaw and McKay (1969) found the lack of social organization and community control in inner-city Chicago to be the primary source of crime. The variables explaining crime must be carefully defined, however, because the differences among areas defy any neat generalization connecting social disorder and criminal activity. Just as in the Dutch example cited in chapter 2, several studies of New York City indicate for each precinct “a strong relationship” between crime and behaviors that lead to misdemeanor summonses or detentions (Kelling and Sousa 2001). But legal scholars such as Harcourt and Ludwig (2006) caution against drawing a direct line between social disorder and crime, because more specific factors such as waves of cocaine use or the presence of certain individuals may have been more significant in New York’s high-crime areas.

Although some scholars such as Gaitán Daza and Díaz Moreno (1994) have found no correlation in Latin America between violence and specific demographic traits such as city size, the instability of urban life worldwide is one common characteristic thought to aggravate crime. As in other regions, urbanization in Latin America usually accompanies economic growth because expanding cities have the jobs and infrastructure that attract and absorb new residents. But if this economic potential fails to materialize or begins to dissipate, cities will struggle to support swelling populations. Latin America reflects the impacts of this phenomenon more clearly than other regions, with an urbanization rate of 75.3 percent, which is comparable with that of far wealthier North America (77.2 percent) and Europe (74.8 percent) and nearly twice the rates for Africa (37.9 percent) and Asia (36.7 percent).9 Some studies in Mexico have revealed how heavy indigenous migration to the cities aggravates youth tensions (Rus and Diego 2007). Such urban pressures are also evident in ownership instability; an estimated 65 percent of Latin America’s urban properties are considered “extralegal,” or lacking valid property rights that would allow their longtime occupants to use them as collateral for loans.10 Although it is applied more widely in high-crime neighborhoods, zero tolerance policing generally does not incorporate either these economic trends or the sociological ones discussed earlier. Limited officer flexibility and discretion, as discussed further in chapter 7, often close one of the few channels for such trends to be brought into policy. For example, a police force that fails to study which properties are targeted by offenders deprives itself of a useful tool for reducing both crime and fear (see St. Jean 2007). Thus geographic distinctions based instead almost entirely on raw crime rates is one reason why zero tolerance is often little more than a mano dura (see chapter 2).

Privatization of Security

Meanwhile, middle- and upper-class areas are more equipped to form neighborhood organizations that can break the cycle of distrust with the police. The national inequality discussed in chapter 2, as it widens the gap in the kinds of residences people can afford, has also made middle- and upper-class areas more physically separate. Indeed, an increasing number of those areas are gated and privately guarded. Many of them have grown from or been based on the self-enclosed communities that have long been part of Latin America’s demographic development, from colonial-era mansions and “company towns” in the 1800s to recreational “clubs” and closed residential districts in the 1900s (Cabrales Barajas 2002; Borsdorf 2003). Amid the modern era’s growth of crime and free markets, security has become a public service from which these areas have increasingly chosen to opt out. Since the 1980s and particularly in the 1990s, the number of private security firms has grown in every region of the world, from 7–9 percent in the industrialized countries to nearly 11 percent in the developing regions (Frigo 2003). Indeed, private security officers now outnumber their public counterparts in cities throughout Latin America, Africa, and Asia. Every Latin American country has experienced such growth,11 particularly in the number of firms guarding gated communities, which have expanded out of the upper-income areas where they began to the less affluent and different types of residential arrangements such as urban high-rises. In Mexico City, for example, about 750 gated communities with roughly fifty thousand mostly middle-class housing units were built between 1990 and 2001 (Parnreiter 2002).

The popularity of the private sector, however, contrasts with its transparency. Because of the constant turnover, off-the-books hiring, and high bankruptcy rates resulting from small profit margins and high-interest loans, the information on private security firms is notoriously unreliable, with the basic statistics varying wildly (see table 3.2). The reported number of private security officers in Mexico is up to 140,000 registered officers and an estimated 600,000 unregistered officers.12 The 2007 estimates for private security officers in Guatemala range from 80,000 to 200,000.13 In total, about 1.6 million registered private security employees are working in Latin America, and about 2 million more are working informally or illegally.14

The effects of private security on both the state and society are far-reaching. Within the state, governments buy into privatization in different forms. For example, the wealthiest of Lima’s municipalities contract out to private firms for policing,15 and the police in most cities say they depend on their private sector colleagues for information. Because many private security agencies are staffed by moonlighting state officers, there is much overlap between the public and private agencies, even when governments do not enter into formal contracts. Laws regulate these enterprises, but, according to estimates by police administrators in most countries, well under half abide by them and so operate essentially unfettered. The lack of oversight in the public police also affects security, with many on-duty state officers prioritizing the businesses and homes they serve off-hours. A bigger danger of such lax regulation is that many employees of the private agencies were dismissed from the force or unable to get in, with obvious implications for both the quality and safety of the services they provide.

Table 3.2 Estimated Size of Latin America’s Private Security Sector

Criminal Policy

Decentralization, private security, gated communities, and other geographic trends can be said to relieve pressure on the state by sharing and dispersing its security burden. But in most cases they also deflate the momentum for criminal justice and other security reform. Privatization also helps to justify casting the net of zero tolerance more widely over “high-crime” areas and the people associated with crime, disorder, and moral decay. Policy then becomes a channel for discrimination against youth, immigrants, indigenous persons, the dark-skinned, sexual minorities, and other sectors associated with criminality and particular neighborhoods (see Benjamin 1969; Foucault 1995; Galtung 2004). As discussed throughout this book, people from these groups who are also newcomers—increasingly common in the flexible labor markets of the neoliberal era—feel distrusted in both residential and commercial areas. As private guards become increasingly ubiquitous, the exclusion upon which their work is based takes on a legal sheen and societal justification. In this environment, legal interpretations by the police, such as the assertions of one Venezuelan state police chief that police codes restrict certain people from moving “freely in the streets,” become unremarkable and unchallenged assertions of police power.16 As a price for the legitimacy that political consensus brings to the law, constitutional rights are not given to all citizens but are instead portioned out and selectively distributed (see Habermas 1998).

When geographic lines spin out and crisscross society, they divide up not just the law and political jurisdictions, but groups and neighborhoods as well. “Apart from examining legal systems and the practices of social control agencies, explanations of exclusion require an account of barriers, prohibitions and constraints on activities from the point of view of the excluded,” points out Sibley (1995, x). In Latin America, where the excluded are a majority confined to a minority of space, such an approach can reveal the hidden nature of citizen security. The exclusivity of consumerist spaces, broad notions of incivility, and long-standing social prejudices all find ready expression in mano dura policies. Such patterns then take on a geographic manifestation as both citizens and state officials divide up the areas in which they live and work as a way to make insecurity more manageable. Thus while individuals associated with crime are “free to go elsewhere but allowed to arrive nowhere” (Papastergiadis 2006), so too are law-abiding citizens restricting the physical ranges of their own lives. Such views may be even more common in the uncertain era of globalization. In areas with diminishing economic autonomy, exclusion of criminal suspects, such as those from certain neighborhoods or deported from the United States, helps maintain boundaries in a world that is rapidly shedding them.

Civil Society

Instead of examining such tendencies, studies of civil society in Latin America and other regions reflect on its role as a catalyst of democracy. Most generally, civil society is regarded as an “arena where manifold social movements … and civic organizations from all classes” organize to “express themselves and advance their interests” (Stepan 1988, 3–4) or to “question” and “resist” the state (Keane 1988, 27–28). For broad “civic culture” theories, civil society is a “school of virtue” that promotes cooperation, aggregates interests, and equalizes representation of otherwise unequal social sectors (Almond and Verba 1963; Putnam, Leonardi, and Nanetti 1994). Narrower conceptions view civil society as temporary constellations of groups that crystallize around particular demands or mediate between the state and the market. Even from these narrower perspectives in which civil society comes to life only on certain occasions, its contributions to democracy are clear. It has mobilized to bring down repressive regimes, forge new policies, and turn itself into a permanent check on state action. From the formation of neighborhood justice centers to the development of community policing, citizen security has been among the many issues to benefit from these efforts.

The democratization literature, in particular, emphasizes civil society. As Latin America threw off its military regimes in the 1980s, this expansive area of scholarship generally began assessing institutional stability, accountability, and effectiveness. As democracies survived but continued to be weak, scholarship shifted to the contrast between democratic durability and vulnerability. Much of this work focused on citizens, pointing out the limited ways in which democratic regimes do and do not incorporate their individual and collective demands, which is one of their primary functions. For example, Von Mettenheim and Malloy (1998, 11) stress civil society’s access to policymaking, and Hagopian and Mainwaring (2005, 8–9) explain how “mass attitudes” and “quality representation” can steer a democracy through rocky periods. Stripping down democracy to the level of individual citizens, O’Donnell, Cullell, and Iazzetta (2004) return to the importance of human agency to revive democracy’s principles by making claims to them.

Although civil society may be the engine of democratization, it is also being altered by the process. This section discusses the connected ways in which civil society directly or indirectly undermines itself and its own security through fear, exclusion, vigilante violence, economic frustration, and responses to youth. As scholars such as Turner (1983) and Habermas (1989) point out, certain conditions or catalyzing events can help civil society overcome constraints to its effectiveness. Incidents of police brutality, for example, can galvanize civil society to push for reform. But unlike socioeconomic or other issue areas in which large sectors of civil society find common ground on identity and goals, the fear of crime and divisions over crime policy often muffle society’s voice and blunt the impact of actions such as protests and local activism. In the process, socioeconomic divisions are reinforced rather than reduced, with zero tolerance leading back to a reflexive resort to physical power. Citizens then easily lose their way in both opinion and impact on their own security. As a fragmented “sphere of publics” rather than “the realm of a single public” (Calhoun and McGowan 1997, 250), civil society thus stumbles in the critical space between awareness and action, and, along with the state, between traditional and problem-oriented policing.17 Thus impediments to specific reforms such as community policing emerge in part out of attributes that are difficult to identify and address through policy.

That said, one of those attributes is fear of violence and crime, which characterizes daily life in contemporary Latin America and fragments its societies. Much work, such as the edited volumes by Londoño, Gaviria, and Guerrero (2000), Morana (2002), and Dammert and Paulsen (2005) on urban violence, explore the impacts of unhealed injustices and the resentments of neoliberal inequality. Going past the “dualistic divide” between academics and policymakers in their approaches to violence, Moser and McIlwaine (2004) describe the conditions that allow vigilantism, gangs, and other manifestations of violence to take root and replicate. Their definition of the causes and attributes of violence has been followed by sociological and anthropological studies of violence that further demonstrate its extent. For example, Briceño-León (2008) discusses the factors that cause violence, such as the gaps between rich and poor and between educational and employment opportunities, as well as the factors that indirectly facilitate it, such as urban density and political crises. Wilson (1998, 311) concludes that “most of the variations among individuals in criminality can be accounted for by personal traits, family socialization, and (perhaps) school influence.” Others distinguish among types of violence. Scheper-Hughes and Bourgois (2004) examine public, private, political, symbolic, and psychic forms of violence, and Buvinic, Morrison, and Shifter (1999) distinguish between goal-centered instrumental violence and “emotional” violence in which the act is “an end in itself.”

But it is social exclusion and individual alienation that perhaps best capture what these disparate strands mean for citizen security and citizen action. As violence becomes predictable in the expectation of its likelihood but unpredictable in its time and form, it contracts the physical and social scope of civil association. As a staple of public discourse and the daily media, violence becomes a part of life to which people pare down their schedules and interactions. The perpetual fear of it reconfigures people’s relationships with each other, with space, and with the state. In many cases, what Rotker (2002) calls a “citizenship of fear” has led to an internal war with fatality rates associated with traditional war but without its markers of ideology and mobilization. As that war’s unwitting combatants, residents learn not to trust “anyone who looks at you for more than a few seconds” and turn to “practical, unpublished maps” with an increasing number of nogo areas (Rotker 2002, 17). Many of those areas are populated by recent migrants from the countryside whose frustration is compounded by the distress of confronting violence worse than that they from which they fled.

The evidence of battle is everywhere. As urban barrios blend into each other, residents retreat into homes barricaded with guns, dogs, walls, and window bars. Around the world, as Young (1999, 18) points out, “the most commonplace barriers and by far the most costly are those that we are forced to erect to protect our own houses.” Many residents in poor areas even say they must have someone home at all times to prevent a robbery, which further depresses household income as well as attendance at neighborhood meetings. Mirroring internal police relations, citizens who are able to bring order to such areas gain reputation and informal power as they organize gangs, private militias, vigilante groups, parapolice squads, and other forms of organized revenge. Such groupings have always existed in Latin America, but they have now adjusted to and have become part of democracy.18 They add to public insecurity, complicate crime fighting, and implicate the state for both its ineffectiveness against and collusion with criminal activity. Many of these groups are formed by police officers, using police uniforms and credentials, and fund themselves by trafficking drugs and contraband vehicles.19

Many ordinary citizens regard vigilantism as an expression of their priority of safety over rights and as their own application of the state’s mano dura policy to achieve that priority. Even when adjusting for documented increases resulting from greater attention to the issue, the number of lynchings and the justifications for them have clearly risen appreciably since the mid-1990s. By most estimates, vigilantism is responsible for the deaths of about six thousand people each year in Latin America.20 Even in relatively low-crime cities such as San José, one-third of the people expressed support for “social cleansing”; in higher-crime cities nearly three-quarters did so (Londoño and Guerrero 1999). This increase is driven in part by the diversification of vigilante groups, which ranges from spontaneous mobs to “social work” groups to clandestine organizations with police ties. Their targets have also widened. The initial focus on individuals suspected of particular crimes has extended to those from marginalized social sectors to police officers and mayors suspected of corruption. Vigilantism is also fueled by customs from the days before democratization. In countries such as Guatemala and Bolivia the practice is most common in areas where non-state or semistate entities, such as local militias, were in control of security during the authoritarian era (see Abrahams 1998; Snodgrass Godoy 2006). Vigilantism also reflects fissures, both avoidable and unavoidable, in Latin American law. As discussed in the section in chapter 2 on penal process codes, the law comprises both rules and the rules about the rules. Even if certain rules are accepted, such as against armed aggression, the rules about applying them may not be. The inefficacy of the state, along with the societal mores that are changing more rapidly than the state, often opens up schisms over how to handle violations of the law.

In Venezuela, where homemade brochures in the early 1990s exhorted barrio residents to “declare war” against delinquency, reports by newspapers, NGOs, and the police reveal an increase in lynchings from only a handful in 1995—a year when 57 percent of respondents in a national poll favored the practice21—to almost one a week since 1997.22 Even in barrios with better physical conditions and community organizations, meetings of “self-defense” groups attract a large attendance, and lynchings have become more open and more brazen, with bodies often left in the middle of the street.23 The two factions of the Tupamaros in the Caracas barrio of 23 de Enero demonstrate how vigilante groups can appropriate law enforcement. The main faction, “Living Hope,” was officially organized to engage in social work and received state funds for it through phantom NGOs. But its members gradually became more involved in crime, both for and against. Many of them have been accused of recruiting young men to sell drugs, as well as to kill over a dozen “presumed delinquents” in the name of its anti-crime campaign. Residents willing to discuss the Tupamaros seem to either fear or tacitly support them. One young resident approvingly described the group’s anticrime tactics. When group members witnessed a robbery, he said, “they tied up the robber, poured gasoline on him, and set fire.” They responded similarly to a man “smoking crack in front of children. Twice they advised him to stop. The third time, they went to his house at night, took him to the roof, and threw him off.”24

As societal antipathy toward constitutional norms segues into active hostility, acceptance of auto-justicia grows and acts of tolerance become signs less of trust than of cease-fires. As Brysk (2000), Armony (2004), and Snodgrass Godoy (2006) describe, this is the flip side of civil society, which challenges the positive association between civic engagement and democracy. The “ties and groups through which people connect to one another and get drawn into community and political affairs” may include groups and forms of organizing that promote violence and intimidation (Skocpol and Fiorina 1999, 2). Determined to stop security from slipping into the uncertainty that characterizes economics and politics, society sacrifices its most vulnerable sectors. As inequalities and tensions multiply, civil society becomes “not an inherently consensual arena” but “a terrain of struggle” over the distribution of and access to resources, entitlements, and, increasingly, rights. As fear of crime spreads further across neighborhoods, times of day, and types of people, it “robs subjects of their capacity to act with or against others. And when large numbers fall under the dark clouds of fear, no sun shines on civil society.… Fear eats the soul of democracy” (Keane 2002, 235).

Despite the depth and extent of violence in Latin America—or perhaps because of it—the region’s governments have not come to grips with it. Even with better statistics and quantification of the economic costs of crime, officials are slow to develop holistic responses to fit the scope of the problem. Beyond specific crimes, this gap between action and the reality of violence is part of the region’s citizen security crisis. Along with its extent, the complexity and local attributes of violence further underscore the need for an approach that accounts for the specific conditions in which violence thrives. Changes over the years in Latin America’s poor urban areas, along with conversations with their residents, put all these factors in stark relief. Killing for tennis shoes and cutting off fingers for rings in Caracas, a frantic robbery of relatives’ homes for cocaine money in Buenos Aires Province, and knife fights among neighbors in El Alto, Bolivia, all reveal the unnerving and unpredictable nature of violence. It is no wonder that policymakers shy away from hitching their careers to antiviolence policies.

A part of this violent push back against insecurity is a push against rights. Although human rights were a catalyst of the antiauthoritarian movements, they have been slowed and sometimes halted in the democracies experiencing growing opposition to rights for presumed criminals. Human rights language and strategies have even been co-opted, such as when the “right” to security is used to trump other rights and to rally mass protests in countries such as Argentina, Mexico, and Guatemala. Perhaps most revealingly, for most Latin Americans “impunity” no longer concerns unpunished abuses from the authoritarian era but unresolved crimes in the democratic one. In much of Latin America, rights and justice have become increasingly contextual—proper in some cases but not others and for some people but not others. Human rights organizations, whose role against dictatorships earned them lasting influence and respect in contemporary democracy, have broadened their focus to include current police practices (see Fuentes 2006a). But in the process they find themselves cast as defenders of criminals rather than of rights, thereby often inadvertently bolstering police resistance when they take the same kinds of critical stances that worked against the military. Such positioning can also signal to executive officials that they can move against individual liberties. In Venezuela, for example, the Chávez government faced few repercussions for rebuffing the rights organizations pressing it for information on police practices. Amid such forms of “rhetorical coercion,” progressive policies and the kinds of citizenship critical to a democratic society are run aground (Krebs and Lobasz 2007, 412). Such practices extend to the daily acts of the criminal justice system. “A situation in which one can vote freely and have one’s vote counted fairly, but cannot expect proper treatment from the police or the courts severely curtails the citizenship” at the core of democracy (O’Donnell 1993, 1361).

Socioeconomic conditions also underscore the need to look past the traditional explanations to understand how civil society is fractured by the conditions in which it lives. With one in three Latin Americans living on less than two dollars a day, many of the root causes of crime are certainly socioeconomic. Substandard education, unemployment, and poverty in general have all been correlated with crime (see Gabaldón 2002, 254). Correlation does not mean causation, however, and so proving a direct link requires specifying the variables in any proposed causal relationship. To demonstrate this relationship, statistical studies of economics and crime often play out first through direct statistical connections, usually with time lags to account for changes over time, followed by a deeper look at the explanatory phenomenon behind the numbers. Focusing on the statistical relations, a study by Fajnzylber, Lederman, and Loayza (2000)—probably the most comprehensive look at straightforward links between crime and socioeconomic determinants in Latin America—finds that the per capita gross national product (GNP) changes signs in different regressions (that is, it has both positive and negative correlations) and that education is not statistically significant. Other studies also indicate that households with low levels of education or income do not report significantly higher levels of violence (Bautista Lara 2006, 45). At the macroeconomic level, low-income countries such as Nicaragua, Peru, and Paraguay have relatively low violent crime rates, whereas several higher-income countries such as Brazil, Mexico, and Venezuela have relatively high ones.

In many studies the other socioeconomic condition most consistently and directly linked to crime is inequality. Affirming a 2002 study by Fajnzylber, Lederman, and Loayza, a report by Di Tella, Galiani, and Schargrodsky (2002, 4) states that “income inequality, measured by the Gini coefficient, has a robust, significant and positive effect on the incidence of violent crimes.” This association continued in work throughout the decade such as in a 2008 study by Briceño-León, Villaveces, and Concha-Eastman of the link between inequality and homicide in Latin America. Ever since data on inequality have been collected, Latin America has had one of the world’s highest levels. With the richest 10 percent having an income eighty-four times higher than the poorest 10 percent (Kliksberg 2001, 106), the region’s overall inequality rate of 49 is much higher than Africa’s rate of 44 and East Asia’s rate of 32 (Casas, Dacha, and Bambas 2001, 24). All Latin American countries have higher coefficients than the global average of 40, and those of several countries reach as high as 60.25 The highest-earning 10 percent of Latin Americans receive 48 percent of national income, while the poorest tenth receive just 1.6 percent, compared with 29.1 percent and 2.5 percent, respectively, in rich countries (de Ferranti et al. 2004). Like the gross domestic product, inequality has been largely regarded as having one of the clearest social impacts as years of stagnation increase frustration over deferred promises of change. But growth has picked up significantly in much of the region since the 1990s, leading to higher and more equally distributed GDPs. Inequality, too, has subsided, with particularly significant decreases in Brazil and Mexico, the region’s two biggest countries. Even as inequality has been pinpointed as having the strongest correlation with crime, many studies, such as Neumayer (2004), show that link to be spurious or insignificant.

Table 3.3 indicates that even a time-based relationship is not straightforward. It correlates change in the homicide rate with three socioeconomic indicators: inequality, unemployment, and the Human Development Index (HDI), which combines life expectancy, literacy, education, and GDP per capita. Placing these statistics on perpendicular axes provides a picture of regional trends that helps visualize the statistical strength of the relationship between homicide and socioeconomic conditions. Figure 3.1 charts homicide against inequality, and figure 3.2 charts it against the HDI. In figure 3.1 the spread of countries over all four sectors, with only a minority in the right-hand sector reflecting parallel patterns, indicates that a decade of change in the homicide rate does not correlate significantly with the economic variable most linked to it. Indeed, the single largest group of countries experienced declining homicides amid rising inequality. In figure 3.2, all countries’ HDIs increased, but less than half had a corresponding decrease in homicide, which runs counter to the assumption that improved HDI decreases murder rates.

Table 3.3 Socioeconomic Conditions: Change, 1995–2005

The intuitive but often statistically weak link between socioeconomic factors and crime points to a need to better specify how and when they are connected. Many studies do so by looking at how poverty incites violence (Concha-Eastman 2002; Salama 2008). For example, finding that GDP growth and the Gini coefficient are statistically significant, the 2000 Fajnzylber report asserts that “stagnant economic activity … induces heightened criminal activity,” and “the fact that this result holds not only for robbery but also for homicide rates may indicate that an important fraction of homicides results from economic motivated crimes that become violent” (p. 246). Other authors describe how inequality has become a symbol of societal antagonism toward free trade, an unstable labor market, and the “plundering of the financial sector” (Reygadas 2006, 127) associated with the neoliberal economic policies adopted throughout Latin America since the 1980s. More specifically, Sánchez, Núñez, and Bourguignon (2003) argue that the poverty-crime correlation is strongest when poverty rates increase rapidly—that is, poverty may foment crime when it heightens frustration and dehumanization in ways that overwhelm mores against violence. In weak states with uncertain policies, low expectations of punishment only intensify that pattern. And if changing citizen attitudes toward legal norms explain how economic problems lead to crime, then they also reveal how much civil society itself has changed as well.

Figure 3.1. Relationship between Changes in Inequality and Homicide Rates, by Country

Source: Homicide rate: see appendix A; inequality: UNDP (1998, 2008).
Note: Horizontal axis indicates the change in homicide rate, measured by murders per 100,000 persons between 1995 and 2005. Vertical axis indicates changes in the ratio of the country’s top 20 percent in national income to the bottom 20 percent in national income between 1995 and 2005.

Figure 3.2. Relationship between Changes in HDI and Homicide Rates, by Country

Source: Homicide rate: see appendix A; HDI: UNDP (1995–2005).
Note: Horizontal axis indicates the change in the homicide rate, measured by murders per 100,000 persons between 1995 and 2005. Vertical axis indicates the changes in the national Human Development Index (HDI), which combines measurements of life expectancy, literacy, education, and GDP per capita, between 1995 and 2005.

Like vigilantism and poverty, youth crime also demonstrates the impacts of citizen security on civil society. Violence by and directed toward youth, at rates that far surpass those of any other age sector in Latin America, is arguably the single most serious threat to the region’s current and future well-being. Violence—homicide, suicide, and accidents—killed eighty thousand Latin America youth in 2006. About 29 percent of all homicides occur among people 10–19 years of age, and fully 36 percent of annual deaths of young men are homicides (Vanderschueren 2007), with the result that the youth homicide rate is up to three times higher than the national rates. At 36.4 killings per 100,000 persons between 10 and 29 years of age, Latin America’s rate far surpasses those of other regions, which range from 0.9 per 100,000 persons in the high-income areas of North America and Europe to 17.6 per 100,000 persons in Africa (WHO 2007). The three countries with the world’s highest youth homicide rates are in Latin America: Brazil, Colombia, and Venezuela.26 Youth unemployment is up to three times the general rate in these and many other countries, and only about half of all Latin Americans finish high school (Kliksberg 2001, 109). Contradictory policies and restricted uses of public space not only contribute to these conditions but also demonstrate how unsettled society is in its response to them.

After the term juvenile delinquency was coined in 1815, its legal definition gradually expanded, along with that of police power (see chapter 2). Originally limited to serious crimes, the term soon encompassed a wide range of antisocial or marginal behaviors. As discussed throughout this book, youth crime has also become a central focus of Latin American policy. According to Alladi Venkatesh and Kassimir (2007, 7–8), the current era of democracy then “accelerated two trends” in legal approaches toward youth “clearly in tension if not outright contradiction”: first, “a punitive turn in juvenile justice and the treatment of young people in public space,” and, second, “an explosion of new legal instruments, discourses, and organization around the rights of children and young people.” Restorative justice and rehabilitation are gaining ground in Latin America as applications of the second approach (see chapter 7). But clashes between these two strategies still dominate policy in the region, especially as a greater number of agencies with different views become involved. For example, the Human Rights Commission of San Luis Province in Argentina has brought greater attention to young people’s socioeconomic rights, since 70 percent of them are not completing high school, making “juvenile delinquency a predominant” source of insecurity throughout the province.27 Many of the resulting policy clashes center on basic legal questions such as minors’ legal responsibility. The United Nations Standard Minimum Rules for the Administration of Juvenile Justice, commonly known as the Beijing Rules, state that legal responsibility “should not be fixed at an age that is too early, taking into consideration the circumstances that accompany emotional and intellectual maturity.” The consensus of experts is that the earliest such age is 15, which is reflected in the laws of most countries. In most of Latin America, juvenile law covers youth between 16 and 18 years of age. They are held responsible for their acts, but in a separate legal category as “minor offenders.” And yet because of the increasing number of arrests in nearly every country of persons between the ages of 14 and 18—for example, the chief of the Chilean Carabineros reports that in Chile arrests in this age range rose five-fold between 1986 and 2003 (Blanco and Bernales forthcoming)—there has been growing pressure to stiffen the penalties against this group and to lower the minimum age for legal responsibility to 14. Problems such as poor education, family violence, and the geographic patterns discussed earlier are quickly overshadowed in such debates.

Meanwhile, these debates feed into broader policy conflicts between crackdowns and prevention. El Salvador, for example, seems unable to move past its contradictory super mano dura crackdowns and its mano amigo social programs and court reforms.28 In Nicaragua, many strong social programs from the 1980s continue to provide generous support for youth, but in 1999 the country enacted the Integrated Development Plan for the Prevention of Juvenile Violence “to clamp down on visible youth crime and thereby restore a sense of security among the urban populace,” and the police systematically arrested suspected gang members without judicial warrants (MacClure and Sotelo 2003, 681). Despite socioeconomic conditions similar to those in its gang-infested neighbors to the north, Nicaragua has kept gang memberships low—and both approaches are credited for that result. Even when minors enjoy relatively strong rights protection, regular police actions toward youth in most countries are far more selective and discretionary than they are for other social sectors. Laws that are less clear and more flexible, such as dispersing crowds after a concert, tend to affect youth disproportionately. For example, the most serious confrontations in Argentina since democratization have involved police and youth between 15 and 18 years of age, and police profiling in the United States excessively affects African American youth. In short, to maintain control and respond to community demands, police divide society into groups that they serve, such as elderly people, and those that they control, such as youth.

That tendency toward control is most apparent in the use of public space. In contrast to adults, who are withdrawing more and more into private or guarded areas, most Latin American youth depend heavily on public space, because many of them are crowded out of their homes and because their identity, self-esteem, and status are built on how they act socially. In the midst of socioeconomic instability, however, these processes of socialization are warped. Generally, the higher levels of vulnerability, the hostility to authority, and the resentment at having adult responsibilities all amplify youth problems in Latin America. When these traits are combined with the consumption of drugs, music, and clothes, the rest of society regards adolescent socialization as unfit for public space. In polls, residents typically identify groups of youth in the street as the neighborhood’s biggest “problem,” whether or not they do anything illegal. Such responses help legitimize even extreme responses, such as killing teenagers in Honduras’s basketball courts. Thus for local police officers striving to fulfill a community’s sense of order, control of youth is almost always a priority. But the skills they need to provide and sustain that order, such as identifying causes of crime and developing responses to them, are the very ones most lacking in their education. Instead, youth policy is painted in broad strokes. Loitering, loud voices, and myriad forms of misconduct—such as “deviance” or “lack of adaptation,” in the words of some police officers interviewed for this book—become characteristics used to mark youth delinquents and serve as a legally backed rationale for detention. Ironically, many young people report that the public focus on such behavior often increases its appeal by giving it an “outsider” or “rebel” association. Such valorization of antisocial status, in turn, deepens society’s perceptions of youth not as individuals but as the instigators of local problems and of jarring cultural changes.

Such perceptions allow even greater problems to fester. One of those problems is youth gangs. Although most officials consider them to be the biggest threat to citizen security in Latin America, in most countries research and policy on gangs are marked by stale information and discordant approaches. Reflecting the dominance of traditional over problem-oriented policing, this clash begins with the policy premise that gangs must be controlled rather than understood. Often missing is the recognition that understanding is the first step toward control. In most countries gangs can be identified and understood through many characteristics: level of establishment (from initial perceptions as a distinct group to physical control of a neighborhood); links (to institutions such as schools); cultural touchstones (from music groups to fascist ideology); various benefits (recognition, protection, solidarity) and costs (danger, alienation, conflict, discrimination by the community); goals (from local dominance to international drug trafficking); and actions taken to realize them (from intimidation to armed violence). Failure to document these traits, along with differences among police officials in the definition of a gang, can result in either exaggeration or underestimation, usually leading to greater police discretion without greater effectiveness. Just as many city police in the United States have focused on isolating pit bulls rather than on how owners train them, police officers tend to profile youth who fit certain descriptions rather than the circumstances that cause that behavior. Also lacking are studies and consistency across time, geographic areas, and government jurisdictions. Clashes between different fields are common as well. Sociological studies that advocate improving the status of youth, for example, often collide with criminologists’ emphasis on first combating violence. The resulting paralysis then allows policy to be hijacked by immediate political needs—a tourism campaign, for example, can turn a group of kids into a gang. Meanwhile, undercurrents are often undetected. In particular, the failure to understand socialization allows the gangs in Central America to create the culture, organization, and language that lay a foundation for growth by appealing to children and meeting their needs for structure, identity, and protection.

Although youth are already the primary targets of policing, societal demands for security around youth are among the tensest and most immediate. They sharpen the dilemma of choosing between traditional and problem-oriented policing, with efforts to address poverty and education clashing with clamors for crackdowns. Across social sectors, families want to protect their own kids while also demanding actions against others. Youth, who are the objectives of policy rather than contributors to it, are reluctant to go to the police even when they are victims—and they are more often victims than victimizers. As a result, most contact between youth and police is characterized by conflict and distrust. Although the continuation of such clashes exposes the drawbacks of traditional policing, amid public pressures for action they also can justify its continuation. Even when such encounters are mediated in public meetings that allow students to describe in detail their complaints about the hostility of residents and businesses, as well as the lack of public space and after-school jobs, the lack of follow-up only adds to their frustration.

Acting on this fragmented template, the civil society that instituted modern democracy in Latin America is not the same one that is trying to maintain it. Fear, exclusion, vigilante violence, economic frustration, and responses to youth may all help keep insecurity at bay in the short term, but they may make it worse in the long run, in part by weakening the capacities of the society that resorts to such approaches.


Because they add to the obstacles outlined in chapter 2, the patterns in power, geography, and civil society just described also affect democratic principles. The supposed trade-off between security and civil rights facilitates special operations, “social control” laws, and other repressive actions that remove criminals from the street but put a squeeze on the rule of law and other democratic standards. The neoliberal globalization of the last twenty years has limited what government can provide just as more questions emerge about whether democracy has lived up to its expectations. Democratization is not a tide that lifts all boats—in Latin America and other regions, agencies and organizations differ widely in their power and their acceptance of constitutional norms. Furthermore, the less police officials have to listen to elected leaders, the weaker is the legitimacy of those officials and the state’s monopoly on legitimate violence. When such conditions lead to a public questioning of democracy or the overthrow of democratic regimes, then the threat that a security crisis poses to democracy is evident.

1. See United Nations Development Programme,

2. By the liberal use of adjectives such as “imperfect” or “illiberal” the literature has done a better job of describing what these regimes are not instead of what they are. See Armony and Schamis (2005), who argue that the literature has created a false division between “old” and “new” democracies.

3. Between the late 1960s and early 1990s the presence of the U.S. Drug Enforcement Agency (DEA) abroad expanded from eight cities to over seventy sites; the number of DEA officials rose from 34 to 416 (Nadelmann 1993, 483).

4. In the State Department the main agency is the Bureau for International Narcotics and Law Enforcement Affairs. In the Justice Department reform is conducted mainly through its International Criminal Investigative Training Assistance Program. And in USAID it is conducted through Governance, Conflict Management (GCM) and the Office of Transitional Initiatives (OTI).

5. This dual definition is used by Schedler, Diamond, and Plattner (1999), as well as by those favoring narrowing it to the state actors responsible for the law. It is also included in discussions of both horizontal accountability among state agencies (O’Don-nell 2003) and the “vertical relationship between citizens and rulers” (see Schmitter 1999, 59; see also Schmitter and Karl 1991).

6. Defensoría del Pueblo, Peru,

7. Mario Zamora Cordero, adviser, La Defensoría de los Habitantes de Costa Rica, interview with author, San José, April 7, 2006.

8. “Deadly Message,” The Economist, July 19, 2008, 47; Foreign Policy Magazine On Line, “The List: Murder Capitals of the World,”

9. Most urbanization has occurred since 1950, when 41 percent of people lived in cities. By comparison, such change has taken twice as long in North America, where the urbanization rate was 53.8 percent in 1925 (Cerrutti and Bertoncello 2003, 4).

10. According to a study of twelve Latin American countries by the IADB and Peru’s Institute of Liberty and Democracy, as quoted in Andrés Oppenheimer, “Latin America’s Underground Economy Keeps Booming,” Miami Herald, October 15, 2006.

11. The growth in private security since the mid-1990s is an estimated 9 percent in Chile and 10 percent in Costa Rica, the region’s safest countries (the rate for Costa Rica is based in part on estimated 50 percent growth since 2002). Although this growth began in the 1980s, most Latin American private security firms are less than ten years old. In the Dominican Republic, for example, forty-seven of the ninety-eight firms registered in 2000 were formed since the mid-1990s (see Díaz 2001).

12. Lower estimate of 140,000: The Economist, January 27, 2007. About 70 percent of private security officers are former state police officers (Uildriks and Tello 2010, 211).

13. Both estimates are a leap from the estimated 15,000 of the 1990s. Lower estimate: Otto Pérez Molina, military general and 2007 presidential candidate, interview by author, New York, June 19, 2007; higher estimate: “Impunity Rules,” The Economist, November 18, 2006, 41. Today the private security force is more than ten times larger than the 20,000-strong police force.

14. Most Latin America firms began in the late 1990s. Although most have fewer than a hundred employees, an estimated 320 have over a thousand (Frigo 2003). In addition, Latin America’s globalized markets allow capture of a large share of the private security sector by multinational firms such as Securitas, Group 4 Securicor, Chubb, and Prosegur.

15. Lima’s municipal security chiefs, interviews by author, Lima, June 20, 2001.

16. Amnesty International, “Urgent Action,” February 12, 2002; Amnesty International (2000).

17. Arendt (1959) describes two actions: to begin (arkhein) and to carry through (prattein).

18. Operating in at least seven states and the Federal District, for example, death squads have been responsible for killings hundreds of Venezuelans. “MIJ evalúa a 327 cuerpos de policía del país,” El Globo, November 25, 2000, 3.

19. For example, one of the very few legal actions against parapolice leaders in Venezuela was paralyzed by the lack of protection for witnesses. “El Defensor no quiere que lo envíen al Rodeo,” Tal Cual, September 18, 2002.

20. Scott Johnson, “Vigilante Justice,” Newsweek International, December 20, 2004.

21. El Nacional, March 14, 1995, and Diario de Caracas, March 15, 1995, 2. That year the interior minister met with leaders of some “auto defense” groups to seek their cooperation “to better guarantee security.” “Escovar Salom se reunirá con empresas de vigilancia y grupos de autodefensa,” Ultimas Noticias, February 2, 1995.

22. See copies of Ultimas Noticias from 1995 to 2000. This newspaper’s crime coverage is both detailed and accurate. Also see PROVEA (2002); El Nacional, March 12, 1997, D-17; and El Informador, June 11, 1998, A-5.

23. Residents of the barrios of Catia, La Vega, and Brisas del Paraíso, interviews by author, 1998 and 2003. “Un hombre linchado en Carabobo,” Diario de Caracas, October 1, 2002, 4–10.

24. Interview by author, Caracas, February 16, 2003.

25. See “Gap between Rich and Poor,”; and IAD (2009).

26. In Venezuela 54 percent of murder victims are under 25 years of age, and homicide is the primary cause of death of poor young people (Sanjuán 2003, 122). In Brazil youth homicides reached 51.7 per 100,000 youth, a rate that is a hundred times greater than that of most countries (UN-Habitat 2004). In the Dominican Republic nearly half of killings are of people between the ages of 20 and 29; 93 percent of El Salvador’s homicide victims are 15–17 years old.

27. Joaquín Sorocco, government minister, interview by author, San Luis, August 8, 2002.

28. On judicial reform, see Call (2003a). For an analysis of the exclusion of youth from legal protection and social services, see Savenije and Van der Borgh (2004).

Additional Information

Related ISBN
MARC Record
Launched on MUSE
Open Access
Creative Commons
Back To Top

This website uses cookies to ensure you get the best experience on our website. Without cookies your experience may not be seamless.